A bug in the electrical system. Noise?

[Murphy625]

Ok.. I'm going to try to stick to just the quick dirty facts.
I have 2 problems or maybe one?

I have built a machine. A programmed hoist to be exact. (60 feet long)

I have one small track section set up in my shop for the hoist carriage to ride back and forth on while I work out problems and polish my programming code. (using Camsoft control equipment).
The horizontal drive for the carriage is a servo motor and the part that lifts the load verticaly is a normal 3ph brake motor driven by a 480volt VFD.
I am using (2) festoon cables which are stacked together. Each is 12 conductors individually screen mesh shielded. All hoist control and power must go threw these 2 festoon cables.
My shop power is single phase 220V. I put this into a 15HP rotary phase converter that I built and the output of this converter goes into a 35KVA 3phase Adjustable Tap transformer.

I am running into what seems to be 2 different problems.

When I test the horizontal travel(servo motor), the hoist ran accurately and trouble free for the 1/2 hour of motion movements that I gave it.
When I test the vertical lifting device (Variable Freq Drive and brake motor), it also works fine.

But when I execute code to have the hoist go to a horizontal position, stop, then lower the lifting device, the VFD trips out on me and gives me a High Voltage Limit error.

The VFD also seems to cause an increasing "Position Error" on the Camsoft interface. I think I solved this however. Still not sure about it but I found a sheild wire I had forgotten to attach on the servo encoder wires. After I fixed it, the position error that occures when the VFD runs is reduced to such a small amount that I dont care. (my hoist only needs +/- 0.5 inches over 60 feet)

Anyhow, here is something I found strange.
If I run the servo, wait for it to completely stop and then kill the servo drive enable circuit before I engage the VFD, the VFD will not always trip out. It seemed like it would only trip half the time.

Anyone have an idea how to solve the problem?
Also, what are your suggestions for seperation space between a 2hp VFD and 1KW Servo drive when mounted in a large panel box? I have mine about 10 inches from each other. Is that enough?


Thanks!
Murphy

 

[DickDV]

As a first step, if the VFD does not have any DC bus snubbing in the form of a small brake chopper and brake resistor, I would add it. Probably should figure on full drive kw capacity for the brake since the sizing is so small.

This will help in two ways. First, if there is regen energy from the hoist down motion, the brake resistor will dissapate the energy. Second, if the incoming AC is noisy for whatever reason, causing high DC bus voltages, the resistor will also absorb that energy.

Little braking packages are not expensive and, in this case, may just be what you need to settle your system down.

 

[Skogsgurra]

Hi, as I said before, you have to change that name ;-)

The VFD Hi DC trip is normal. Lowering the load means regeneration and regeneration means increased DC link voltage - you need to add a braking resistor. There is probably a brake chopper in the VFD already. Or does the VFD have built-in regeneration? Very few have that.

The second question - have no good answer.

The third question (distance between servo and VFD). If the servo is a PWM type it is a heavy EMI generator and as such it can take a lot of external EMI. The same is true for the VFD. So you should be able to mount them close together. Be careful with screen connectios, no long pig-tails. Best is to use clamps that connect "360 degrees" directly to unpainted mounting plate. EMC glands for cable entry and exit is recommended. Yes, it is difficult if you are working with flat cables - but try to get as close as possible to a low impedance connection between screen an mounting plate.

And change that name! Humpry is almost identical (same letters, just move them around a little) and has no law associated with it - not as far as I know.

 

[Murphy625]

First,
Thanks for all the helpful information.

Both replies suggest that the hoist is generating power on its way down. I have limited experience on this so I could be wrong but wouldn't the load need to actually pull the hoist down in order to generate this feedback power?
The lifting device is a 1-1/2 hp motor attached to a 50:1 wormgear with dual output shafts.
If I manually release the motor brake, there is to much resistance in the wormgear drive to even move the shaft.
In other words, I don't think gravity is affecting the downward motion very much. I'm sure gravity makes lowering the load easier but my problem is happening with no real load on the device anyhow.
The dirty incoming power comment gets my attention however.
Time for a braking resistor to soak up some of those spikes yes?

YOU SAID Best is to use clamps that connect "360 degrees"

What are you talking about here? A 360 deg clamp for shield wires?? Please give more info on this.
I do have pigtails!!!!!!! and some of them are 6 inches long. I almost fell off my chair when I read that.

Where should I go for a braking package? Do I call up Reliance electric (maker of the VFD) ??

And the name STAYS... LOL..
I don't mind problems. I just adds to the brain power when I solve them. If you keep mixing up the letters in my name, your going to give me identity problems...

Thank you DickDV and Skogsgurra

 

[Skogsgurra]

Murphy,

A six inch pig-tail is bad. There is a rule-of-thumb saying that a pigtail longer than 1 inch is open circuit at the frequencies that are active here (easily 10 - 100 MHz). If you google for "EMC glands" you will get some 100 hits telling you how to do that part correctly.

You are absolutely right about the worm gear. You shouldn't have any regeneration there. I think that you need to hook up a recorder (Careful! Hi voltage. Use a differential probe or isolation amplifier) and study the DC link. That will guide you to the source of the problem - I hope.

 

[smorein]

A coax shielded signal, with a pigtail connection to ground can be looked at as the following:

A capacitor from the signal source to the shield,
an inductor from the shield to ground.

This is, of course, one way of wiring an antenna (with the shield as the antenna): a capacitor to isolate any dc ofset from the poweramp, and an inductor to ground (blocking AC)

After reading the above (cannot remember where) I stopped ever using pigtails.

 

[DickDV]

Murphy, it is best to buy the braking package from the VFD manufacturer.

 

[Murphy625]

I am starting to think that I need to improve my technique when I attach shielded connections.

But how?
Here is what I have done.

The servo drive and VFD drive output lines are both run with the same type of cable. The cable is 4 twisted 12ga wires with a spiral wound mesh type screen. Both lines exit the panel box by way of a single 1 inch metallic conduit.
This same conduit also contains 2 other single conductors (thhn or thwn wire). These 2 wires provide the power to the motor brake. (480 volts but only 2 of the three phases are used)
The servo encoder wires are run by way of a cable that contains 4 twisted pairs. Each pair is shielded with foil, and then the entire cable has overall shield.

All cables above terminate into the same 12 x 12 junction box which contains a terminal strip with 24 connections. (1 for each festoon cable conductor)

I arraigned the connections so all the high voltage was on the far left side of the box, and all low voltage on the right side of the box.

When a shielded cable enters the box from the EMT conduit, the cable runs all the way up to the terminal strip with the shield intact. The cable was cut long and then the power conductors trimmed back so only about 2 inches of the conductor was left unshielded. This left short power conductors, but a long shield pigtail. Am I to understand that there is a much better way to do this? All of the "EMC glands" that I found seem to terminate the shielding at the enclosure wall. I ran my shielding all the way up to the power terminal connections.

What am I failing to understand here?

Are you guys suggesting that all my power terminals and encoder wires need to be shielded like a serial port cable that terminates into a 9 or 15 pin D-sub connector?

Thanks for all the help..

Oh, and an interesting note.
When I was first installing all this, I turn on the VFD and tested the motor without any shielding(20+ feet of normal thhn wire run in the air). In a panel box located 6 feet away from the VFD (the door was open), I had OPTO22 relays that were flickering on and off without being told to do so. Some of them were even lighting up the indicator lights on the front of the box.

Murph